MIT Study: Machine Learning Models Lack Coherent Values

Artificial intelligence (AI) is advancing rapidly, prompting discussions about whether machines are developing their own moral compasses. A new MIT study challenges these sensational notions by revealing that current AI models don’t possess coherent values akin to humans. Instead of holding ingrained principles, these models function as sophisticated imitators of human behavior based on patterns and prompts.